CONCEPTUAL MARKET ORGANISATIONS FOR THE PROVISION OF INNOVATIVE SYSTEM SERVICES: ROLE MODELS,

ASSOCIATED MARKET DESIGNS AND REGULATORY FRAMEWORKS

WEBINAR 3.2

June 9th, 2020 This project has received funding from the European Union’s Horizon 2020 research and innovation programmeunder grant agreement No 773505.

Agenda

• Context and Overview of EU-SysFlex

• Introduction of WP3 and the Speakers

• Overview of Task 3.2

• Market versus regulated approach products

• Grid constraint management

• Joint procurement

• Conclusions

Webinar Task 3.2 09/06/20202

https://eu-sysflex.com/documents/

Wind & Solar

Natural Gas

Coal

Other Renewables

Nuclear

0%

5%

10%

15%

20%

25%

30%

35%

40%

45%

2015 2020 2025 2030 2035 2040

Share of electricity by source European Union 2017 -2040Source IEA 2018

2017

Today

Future system will be heavily reliant on non synchronous sources of electricity

Context

3 Webinar Task 3.2 09/06/2020

Decarbonisation by 2050

Context

Webinar Task 3.2 09/06/20204

EU-SysFlex Project Structure

Webinar Task 3.2 09/06/20205

EU-SysFlex Project Structure

Webinar Task 3.2 09/06/20206

WP3 deliverables

Webinar Task 3.2 09/06/20207

https://eu-sysflex.com/documents/Webinar #416/6/2020

Presenters

• Przemyslaw Kacprzak, PSE

Department of International Cooperation, PSE

• Niamh Delaney, EirGrid Innovation

principal market analyst

• Philippe Loevenbruck, EDF

Research engineer in power system economics, EDF

• Jan Budke, innogy

Expert for European regulation and projects

• Sylvie Magois, EDF

Research project manager in regulation and market design & task leader

Webinar Task 3.2 09/06/20208

TASK 3.2 OVERVIEW

Webinar Task 3.2 09/06/2020

Objectives of Task 3.2

• To build upon the developed generic role models to describe the responsibilities and interactions between system operators (regulated players) and deregulated players (in particular flexibility service providers), for system service provision by both centralised and decentralised energy resources (demand response, storage, generation).

• To assess different procurement options of generic products necessary to solve scarcities identified in decarbonised and decentralised systems.

• To compare different market/regulatory organisations with existing market designs and regulation in EU and highlight their advantages/drawbacks and the issues raised

• To identify key attention points in the market/regulation options for further investigation supported by advanced quantitative power system and market modelling (Task 3.4).

Webinar Task 3.2 09/06/202010

Overview of the approaches utilised in Task 3.2

Webinar Task 3.2 09/06/202011

(adapted from EU-SysFlex Project, 2018b)

Main Steps

• Adaptation of T3.1 and T3.3 results to our needs : generic roles list, list of generic products, first list of market organisations to be assessed

• Description of interactions between actors for the procurement of the generic products in different organisations (market-based/regulated) with different optimisation’s methodologies

• Assessment of advantages and drawbacks of regulated vs market-based procurement

• Grid constraint management issues

• Joint procurement of several products

Considered services : frequency control, inertia, voltage control, congestion management

Webinar Task 3.2 09/06/202012

MARKET VERSUS REGULATED APPROACH FOR THE

DIFFERENT PRODUCTS

Webinar Task 3.2 09/06/2020

Market-based vs. regulated organization

• Market based approach• Supply meets demand• Equilibrium price

• Regulated approach• Basically „lack of the market”• Regulator regulated the price of service• Could be just interim solution or long-term solution

• Markets may have (and usually have) some regulated aspects

Webinar Task 3.2 09/06/202014

Market-based vs. regulated organization

Webinar Task 3.2 09/06/202015

Parameters characterising the market

Webinar Task 3.2 09/06/202016

Characteristics Description / Options

Market pre-qualification

Nature of the participantsMandatory participation? Based on which characteristics (generator/DSR, location, flexibility size, …)?

Procurement

PerimeterThe area the marketplace encompasses: Local (= DSO level), national or zonal (= one TSO level), cross-border (= cross TSO level).

Frequency For instance, annual, daily, hourly, or even shorter

Nature of the buyerWho is responsible for obtaining a specific product: is the TSO or the DSO the only buyer (monopsony, single buyer)? Or are SOs both buyers? Or even commercial market participants?

Product structure

Characteristics of the product One or several products procured on the same market: 1. one product, 2. several products co-optimised (for instance upward / downward / symmetric or capacity / energy)

Spatial resolution of the product (= location)

Which precision is required for the location of the product?

Temporal resolution The length of the time period in which a specific product is defined

Delivery horizon The length of the quantum of time in which a specific product shall be delivered

Activation

Activation of the productProcedure for the activation of a product: inherent, automatic, manual. Coordination between TSOs and/or DSOs, if required.

Settlement

Verification Rules for verification of actual delivery, including definition of the baseline.

Payment

- Regulated price?- Regulated bid price (for instance, market parties have to offer prices based on their

variable costs)?- Existence of price caps/floors?- Pay as cleared or pay as bid

Penalties In case of non-delivery or non-conform delivery

Capacity vs Energy products

Webinar Task 3.2 09/06/202017

Energy product:

• Energy (active or reactive) used by TSO and/or DSO to insure secure network operation

• Provided by Flexibility Provider

Capacity product:• Capacity reserved (MW or MVAr) that

flexibility service provider agreed to hold and provide required energy if required

• Pricing of activation• Separate biding of energy product (e.g. aFRR,

mFRR, RR), or• Fixed price in capacity bid• Fixed price (including price of zero)

*including delivery of reactive power

Services with only capacityproduct:• FCR, FFR

Services with both capacityand energy product:• aFRR, RR

Services with only energyproduct:• In some countries: mFRR,

redispatching• Wholesale energy in SDAC, SIDC

Four Phases of Product Acquisition

Webinar Task 3.2 09/06/202018

• Regardless of the Market Organisation, there are 4 main phases in product acquisition:

Prequalification:certification (technical, financial and data communication) and registration of flexibility service providers (FSPs)

Procurement:Market bidding, clearing, selection (market) or resource selection (regulated)

Activation:automatic or manual triggering use of flexibility

Settlement:measurement; comparison with expected response; payments calculation; data exchange (e.g. for imbalance settlement)

When to use Market-based, when to use Regulated

Webinar Task 3.2 09/06/202019

• Criteria considered:

• Compliance with EU rules

• Liquidity

• Strategic Gaming and Market Power issues

• Efficiency in short-term Allocation of Resources

• Suitability for Investment

• Simplicity

• Transition Costs

When to use Market-based, when to use Regulated

Webinar Task 3.2 09/06/202020

• Frequency Control Products:

• The Clean Energy Package (Art. 6) and EGBL (Art. 32) require balancing capacity to be procured in a market-based way unless there is insufficient competition• Procurement timeframes also stipulated

• RfG Network Code specifies minimum mandatory capability for generators above a certain threshold

• A market-based organisation should apply to frequency control products

When to use Market-based, when to use Regulated

Webinar Task 3.2 09/06/202021

• Inertia:

• While market-based solutions are preferred, a regulated approach could be used, if necessary, to ease the transition phase (decommissioning of synchronous generators)

• Voltage Control Products:

• Strong mandatory requirements exist in the European network codes regarding voltage control capabilities

• Choice may depend on the nature of the voltage problem:• Voltage problems due to too high feed-in/consumption should be regulated

(mandatory participation, part of connection agreements), as problem can only be solved locally

• Voltage problems due to power flows in transmission/distribution lines far away from their foreseen operating range: a market for daily procurement could be considered as more FSPs can solve issue

When to use Market-based, when to use Regulated

Webinar Task 3.2 09/06/202022

• Congestion Management Products

• Market-based solutions should be implemented in all cases when market power and increase/decrease gaming can be limited

• auctions to procure new capacities with long term agreements

• market-based organisation for short term allocation

• If the liquidity is very poor and increase/decrease gaming cannot be sufficiently limited, mandatory participation with cost-based remuneration for generation and voluntary non-firm connection agreements for loads could be feasible options

General Considerations to Facilitate Participation of FSPs

Webinar Task 3.2 09/06/202023

FSP open participation

Product characteristics

Aggregation rules

FSP revenue

Easy data exchange

Suitable pricing

mechanism

Summary

Webinar Task 3.2 09/06/202024

• There are various mechanisms for the procurement of flexibility services:• Some mechanisms are market-based

• Others are regulatory approaches

• For power systems with a high level of variable RES, market-based approaches are generally preferred • However the regulated approach is still needed – in some cases as an

interim solution, in others as a longer-term solution

• Appropriate design is needed for the efficient development of flexibility services markets

GRID CONSTRAINTS MANAGEMENT

25Webinar Task 3.2 09/06/2020

Main options for consideration of grid constraints during the technical bid selection

• The Optimisation Operator (OO) is responsible for the optimisation of bid selection and switching measures to solve grid constraints and satisfy reserve needs

• How to consider grid constraints (current, voltage) during the bid selection process ?

• Main choice: centralised or decentralised optimisation

• Other choice: which grid data are sent by the SO to the OO?• Comprehensive

• Partial

• Bid limitations

Webinar Task 3.2 09/06/202026

Centralised optimisation with comprehensive grid data

• There exist variants with:• Partial grid data (sensitivities, e.g. for only one topology)

• Bid limitations

Webinar Task 3.2 09/06/202027

TS_O DS_O OO FSP

Send bids selection

Send comprehensive grid data

Assess grid constraint needs via grid

assessment, select bidsand switching measures

Send comprehensive grid data and reserve

needs*

*other needs than grid constraints

MO

Send all results

Validate bids

Send bids selection

Submit bids

TS_O: Transmission System Operator

DS_O: Distribution System Operator

OO: Optimisation Operator

MO: Market Operator

FSP: Flexibility provider

Decentralised optimisation with comprehensive grid data

Webinar Task 3.2 09/06/202028

OO_TTS_ O DS_O MOOO_D FSP

Each OO runs grid assessment and selects

bids for own needs. Coordination between

OO_D and OO_T to find most optimal solution.

Send bid selection

Send bid selection

Send comprehensive

grid data

Send bid selection

Validate bids

*other needs than grid constraints

Send comprehensive grid data and reserve

needs*

Submit bids

Note: decentralised optimisation does not reduce liquidity by design.

TS_O: Transmission System Operator

DS_O: Distribution System Operator

OO_T: Optimisation Operator in Transmission

OO_D: Optimisation Operator in Distribution

MO: Market Operator

FSP: Flexibility provider

Discussion of advantages of centralised and decentralised optimisation

Webinar Task 3.2 09/06/202029

Advantages of centralised optimisation

Advantages of decentralised optimisation

•Less coordination effort between roles needed (by definition of centralised optimisation with comprehensive grid data, there is no iteration in the operational timeline)

•Can achieve theoretical possibility of (fully) optimal solution

•Economy of scale (one vs multiple places for the optimisation algorithm)

•Interoperability concentrates on interface to one OO

•Stepwise optimisation implementation along the voltage levels is possible, considering specific voltage level and regional requirements

•Easier to match localised solutions to scarcities, since no new optimisation of the whole system is necessary

•Simpler individual algorithm, less data processed

•Fit to current SOs responsibility framework and regulation framework

•Higher resilience

To which actor to allocate the OO role?

• Studied for the actors:• Each DSO and TSO • TSOs/DSOs joint venture• Commercial third party

• And for both centralised and decentralised optimisation

• All solutions are feasible and, to properly allocate the roles of OO, it’s necessary to conduct a cost-benefit analysis, considering all chances and risks, but specifically also addressing national specificities (regulation, number of DSOs and TSOs within a bidding zone, existing processes of optimisation, historical organisation, etc.) and choice for centralised or decentralised optimisation.

• Regardless of the national situation, an allocation of the optimisation to an actor other than each individual system operator, being responsible for the safety of their systems under the Electricity Directive, leads to significant governance and regulation challenges. • This report only reveals first implications and further research is needed in

this direction.

• See the comprehensive table in the report

Webinar Task 3.2 09/06/202030

Summary

• Both centralized and decentralized optimisation might be applied for selecting flexibilities

• Both methods are feasible with one or several marketplaces and neither centralized nor decentralized optimisation reduce liquidity by design

• Conceptually, centralized optimization leads to an optimal allocation, but decentralized optimization can get to the same or result or close to it – depending on circumstances and design

• Decentralized: more relevant where DSOs need locational products

• For inertia/FCR: centralized more relevant

• The optimisation operator determines the optimal combination of switching measures and flexibility selection• The allocation of this role can theoretically be allocated to other actors

than each individual DSO/TSO, but this would cause significant governance and regulation challenges

Webinar Task 3.2 09/06/202031

JOINT PROCUREMENT

32Webinar Task 3.2 09/06/2020

No single definition exists, many forms of joint procurement possible

Webinar Task 3.2 09/06/202033

Number of

products

Number of

scarcities

Number of

buyers

Joint

ProcurementOptimisation across scarcities Example

1 1 1 No

No

(separate optimisation of

scarcities)

Separate procurement of mFRR and CM by TSOs; procurement of CM by

DSOs

1 1 2 Yes

No

(separate optimisation of

scarcities)

Coordinated procurement of CM by TSOs and DSOs

1 2 1 YesYes

(joint or coordinated)

Procurement of mFRR type of product for CM and imbalances by TSOs;

Procurement of CM product for CM and voltage control by DSOs

1 2 2 YesYes

(joint or coordinated)

Procurement of mFRR type of product for CM and imbalances by DSOs and

TSOs

2 1 1 Yes

No

(separate optimisation of

scarcities)

Procurement of reactive power and active power for voltage control by TSOs

or DSOs

2 2 1 Yes Yes (joint or coordinated)Procurement of reactive power and active power for CM and voltage control

by TSOs or DSOs

2 2 2 YesYes

(joint or coordinated)

Procurement of reactive power and active power for CM and voltage control

by DSOs and TSOs

Motivation for joint procurement of congestion management and balancing/mFRR

• Recently discussed in TSO/DSO report on active system management (CEDEC et al. 2019).

• Implemented in some countries (e.g. FR, UK)

• mFRR and CM products have similar characteristics (active power, manual activation, FAT and duration ≥ 12.5 min).

• Potential advantages (depending on the version)• joint product could allow FSPs to bid only once (e.g. appropriate in FR,

where few congestions exist)

• joint optimisation could decrease flexibility volume needed (e.g. appropriate in GER, where the likelihood for synergies is high due to many congestions)

• Report also describes disadvantages based on different versions.

Webinar Task 3.2 09/06/202034

CM: congestion managementmFRR: manual Frequency Restoration ReserveFSP: Flexibility Service ProviderFAT: Full Activation Time

Joint optimisation: creation of synergies

Webinar Task 3.2 09/06/202035

Prerequisites for synergies

• joint product

• overlapping need in time and volume

• opposite direction of needs in specific area (here: load pocket)

• optimisation within mFRR energy selection timeframe

German case study

no clear net advantage, higher costs possible due to likely substitution of cheaper (regulated) redispatch bids by more expensive mFRR bids

CM: congestion managementmFRR: manual Frequency Restoration ReserveFSP: Flexibility Service Provider

Product design

• Focus on procurement of energy product near real time• potential synergies of energy product due to joint optimisation will be considered

in capacity procurement process

Webinar Task 3.2 09/06/202036

ToTo-5

flexibility provision

To+15

• Joint product is a prerequisite and must comply with mFRR and CM requirements

• FAT ≤ 12.5 min→ certain FSPs with longer FAT/preparation time will be excluded (e.g. cold generation, storage flexibilities (electrical, virtual such as demand shifting))• locational bids• describe rebound behaviour (if existing), e.g. via max. shifting time

FAT: Full Activation TimeGCT: Gate Closure TimeMARI: Manually Activated Reserves Initiative

Time in minutes

CM: congestion managementmFRR: manual Frequency Restoration ReserveFSP: Flexibility Service Provider

Process design

Webinar Task 3.2 09/06/202037

Version 1: coordinated optimisation via connected bidding

Version 2: coordinated optimisation via joint bidding

Version 3: joint bidding and joint optimisation

1. bidding for CM2. clearing CM3. bidding mFRR

(incl. voluntary transfer of bids)

4. clearing MARI

1. bidding for CM/mFRR

2. clearing CM3. clearing MARI

1. bidding for CM/mFRR

2. clearing CM (consider mFRR needs)

3. clearing MARI

To

GCT for mFRR bids FP ➔ TSO

To-12

flexibility provision

To-7,5To-25

GCT for bids ➔MARI

Time in minutes

MARI clearing and results

GCT CM TSO/DSO

To-Y (≥ 55)

Clearing(s) CM TSO/DSO

To-X

≥ 30 min

Results Results

To

GCT for mFRR/CM bids FP ➔ TSO/DSO

To-12

flexibility provision

To-7,5To-25

GCT for bids ➔MARI

Time in minutes

MARI clearing and results

To-X Results

Clearing(s) CM TSO/DSO

To

GCT for mFRR/CM bids FP ➔ TSO/DSO

To-12

flexibility provision

To-7,5To-25

GCT for bids ➔MARI

MARI clearing and results

To-X Results

Clearing(s) CM + mFRRTSO/DSO

Time in minutes

Comparison of versions: marginal costs

Webinar Task 3.2 09/06/202038

Marginal costs of joint procurement based on a joint product compared to

separate procurement of CM and mFRR

Version 1:

coordinated

optimisation via

connected

bidding

Version 2:

coordinated

optimisation via

joint bidding

Version 3:

joint bidding

and

optimisation

Cost

decreasing

effects

Offering energy flexibilities both for mFRR and CM (one product):

Liquidity increasing effect for the buyers and effort decreasing effect

for FSPs

X

(on voluntary

basis)

X X

Lower flexibility volume needed due to the use of flexibility to solve

both mFRR and CMX

Limited exposure to the risk of strategic bidding X X

Cost

increasing

effects

Break-down of aggregated portfolio to locational bids and – in case of

rebound effects – delivery of max. shifting times lead to an increased

risk margin and increased efforts for aggregators normally bidding

only for mFRR

X X

For CM: Liquidity decreasing effect due to exclusion of FSP with

preparation times non-capable for mFRRX X

Exclusion of some grid flexibility potential (shifting grid

maintenance measures to allow topology switching) in the

optimisation of CM measures

X X

For CM: Exclusion of FSPs which can only place bids in the

timeframe of 25-55 min before delivery or increased risk margin in

case of (allowed) aggregation

X

Comparison of versions: challenges

Webinar Task 3.2 09/06/202039

Challenges of joint procurement compared

to separate procurement of CM and

balancing

Version 1:

coordinated

optimisation via

connected bidding

Version 2:

coordinated

optimisation via joint

bidding

Version 3:

joint bidding and

optimisation

Timeframe currently not sufficient for CM

optimisation, incl. TSO/TSO - DSO/DSO

coordination

X X

No possibility to allow congestion-specific

solutions (with regard to local marketplace

design, product design, bidding and

optimisation across several time intervals,

etc.)

X X

Higher complexity of the algorithm(s)

and/or coordination between SOsX

Cost allocation to SOs for CM and BRPs for

balancing (see Electricity Regulation, Art. 13) X

Transition costs Depends on already implemented national solution

Summary

• JP of mFRR and CM energy product is a relevant option.

• Quantitative analysis needed to determine whether JP and which specific version is superior to separate procurement.

• German country case study (2018) did not reveal convincing benefits.

• To be considered on national basis• trade-off between potentially decreased costs and increased complexity

depends• severity of grid congestions and mFRR needs• cost structures of mFRR and CM• existing processes

• other options• hybrid approaches of joint/separate procurement• where RR is existing: joint procurement of RR and CM

Webinar Task 3.2 09/06/202040

CONCLUSION AND NEXT STEPS

41Webinar Task 3.2 09/06/2020

What we did

different procurement arrangements of generic flexibility serviceshave been studied by Task 3.2 and feasibility, advantages and drawbacks of different options have been assessed:

• regulated organisation compared to a market-based organisation

• methodologies to consider grid constraints in the flexibility procurement : in particular centralised optimisation of flexibilities compared to decentralised optimisation

• allocation of the optimisation operator (OO) role to each individual system operator or other actors

• joint procurement of services, in particular for mFRR and congestion management

Webinar Task 3.2 09/06/202042

quantitative studies needed to conclude

Quantitative studies are needed to evaluate the proposed solutions

• Efficiency

• Investment or transition costs for the implementation of the new market design as well as operational costs for running the optimisation and the marketplace

• Temporal and computational feasibility

• Communication costs

• The allocation of the OO role to system operators or other actors and the type and amount of grid data to be exchanged between actors.

• The question of small DSOs

Webinar Task 3.2 09/06/202043

FSP’S REVENUE ISSUE

Would FSP revenues be sufficient to ensure an appropriate deployment of flexibilities ?

• the need for flexibility and capability to deliver specific system services will increase (T 2.1).

• Task 2.5 has demonstrated that an energy only market will not provide sufficient revenue to cover investment costs => issue regarding the volume of flexibility available

• Task 3.4 reaches similar conclusion

Task 3.2 describes some market design solutions, e.g. the proposed long-term capacity product that can foster investment in new flexibilities if short-term markets are not able to provide these signals.

It is of great importance to properly quantify revenue streams available for all sources of flexibility in order to adapt market design proposals adequately.

Webinar Task 3.2 09/06/202044

This project has received funding from the European Union’s Horizon 2020 research andinnovation program under grant agreement No 773505.

Thank You!

Questions?